Deep Learning Techniques for Sentiment Analysis

Abstract

Sentiment analysis covers a wide range of computational research, including research on the opinions, feelings, emotions, evaluations of people, and attitudes toward products, services, organizations, individuals, issues, events, topics, and their attributes. It plays an increasingly important role in the era of big data. In fact, it has spread from computer science to management and social sciences such as marketing, finance, political science, communications, medical science and even history, generating common interest throughout society due to its commercial importance. TEA is a basic task with the typical used of NLP methods which is full of interest, particularly for fine-grained classification of textual emotional content. It is the process of mastering, inductive analysis and reasoning about emotional content. Simply put, it is the process of analysing, processing, summarising and reasoning about emotive and subjective texts. The Internet generates a large numberof user reviews to gain valuable information about people, events and products. These reviews express a wide range of emotions and emotional tendencies, including joy, anger, sadness, delight, criticism and praise. Potential users can therefore view these subjective reviews to understand how public opinion views an event or product.